Lisa and I bought the first Macintosh [128K RAM, 400K floppy disk] at release in the Spring of 1984 through the Apple University Consortium program at Northwestern University (I was working full-time at the Vogelback Computing Center and taking 2 classes/quarter for my MS in CS; Lisa was working on her BA in CS full-time). I had to write up my master’s project report, the beta version of MacWord crashed, so I went to nabih’s in Evanston and bought a copy of Microsoft Word for Macintosh. [nabih’s opened in 1972, started carrying Apple products in 1978, and are still in business today.]
Having worked at IBM from 6/1982 to 6/1983, I knew the mainframe computer (and IBM) was dead, the minicomputer (e.g., VAX) would die next, and the future was microcomputers! Apple never answered my snail mail letter, but my 12/5/1984 letter to Microsoft yielded a snail mail reply on 12/17/1985 (dated 12/13) from Jo Ann Rahal, Recruiter. She asked for code samples!
I quickly wrote up a cover letter and printed out four code samples and mailed them Bellevue.
PPSTAT — a program to collect statistics on CDC-6400 peripheral processor program usage.
PANEL — a full-screen GUI library for DCL on the VAX 11/780.
WSCONV — a short C program to convert WordStar files to ASCII text.
SIMULATE — a large operating system simulator for a class, written in Pascal (w/Lisa).
Ms. Rahal sent me a Western Union Mailgram (telex) on 1/9/1985 at 18:45 PST and it was printed out at 21:52 EST (a 7 minute journey). It asked me to call collect, I did (I presume on 1/10) and we agreed I would depart Chicago on 1/31 at 7:04p and depart Seattle at 9:40a on 2/3.
I woke up early on Friday 2/1 to discover fresh snowfall! Something Seattleites are not used to! But I had been driving in the snow in Poughkeepsie, NY, and Evanston, IL, for three winters, so I cruised in my front-wheel-drive Avis rental car from my parents home in Seattle to the Northup Way Microsoft building in Bellevue.

My memory of my interview day is quite limited forty year on, but I recall Gordon Letwin (#8 hire) was my “as-appropriate” interviewer. I had done a lot of COMPASS (assembly language) programming on the CDC-6400 at Northwestern, so he asked me how I would implement a stack pointer for call/return (instructions missing in the CDC CPU). I remember fumbling at first, but coming up with something workable.
I returned to Ms. Rahal’s office, and she gave me a job offer: $31,000 salary, two weeks of vacation, and options to buy 1,500 shares of stock at $3/share. She said she thought those options might be worth $20K-$30K some day. I had no clue what a stock option was! I liked the people I met, and $31K was a nice bump from the ~$23K I was earning at Northwestern.
I got back to Evanston and received a snail mail letter dated 2/4/1985 from Steven A. Ballmer, Vice President, Systems Software. With a “cc: Bill Gates”. It confirmed the verbal offer, adding a bonus of up to 15% of my salary, medical and dental insurance, relocation expenses, and membership in the P.R.O. Racquet and Health Club. I send a short acceptance snail mail letter back on 2/9 (I assume the day I received the printed offer letter).
A few days later, I received another Western Union Mailgram :
1-0194771092 02/21/85 TLX MICROSOFT BVLE CGBA
24 BELLEVUE, WA, 13:54 PST 02/21/85
BENJAMIN W. SLIVKA
VOGELBACK COMPUTING CENTER
2129 SHERIDAN ROAD
EVANSTON, IL 60201
I AM HAPPY TO HEAR YOU HAVE ACCEPTED OUR OFFER TO JOIN MICROSOFT. WE BELIEVE GOOD PEOPLE LIKE WORKING WITH OTHERS WHO ARE AS CAPABLE, ENERGETIC, AND DEDICATED AS THEY ARE. PART OF THE UNIQUENESS WE HAVE IS BECAUSE OF OUR CARE IN SELECTING PEOPLE TO JOIN MICROSOFT.
WE ARE PARTICULARLY IMPRESSED BY YOUR QUALIFICATIONS. OUR PROJECT TEAMS ARE QUITE SMALL AND RELY ON A STRONG CONTRIBUTION BY EVERYONE. WE WILL PROVIDE A CHALLENGING PROJECT AS SOON AS YOU ARRIVE.
I LOOK FORWARD TO SEEING YOU HERE AT MICROSOFT.
BEST REGARDS,
WILLIAM H GATES
CHAIRMAN OF THE BOARD
TLX328945
*G/BMS
In April, Microsoft arranged for a house hunting trip. I flew out on 6/1 and Lisa joined me on 6/3. Our return flight to Chicago was on 6/9. We found a small apartment in Kirkland that was a 2-mile drive from the Northup Way building.
Using the paper maps available in 1985, I mapped out a six-day, 2,089 mile drive from Evanston, IL, to Bellevue, WA. We drove through Wisconsin, Minnesota, South Dakota, Wyoming, Montana, and Idaho on the I-90. [Google Maps today claims that is a 2,039 mile drive.]

I arrived at 10700 Northup Way, Bellevue, WA, early on Monday 6/24/1985. I was given an IBM PC-AT: an Intel 80286 at 6 MHz, 1 Mb of RAM, 20 Mb hard drive, and an EGA card and color CRT display. There was no LAN (let alone Wi-Fi): we did source control on a DEC PDP-10(?) mainframe running Xenix, using FTP to upload and download files over a 19 kbps hard-wired serial line. My first project was MT-DOS (Multi-Tasking Disk Operating System). Our goal was to run multiple MS-DOS apps in 640K of RAM! Anthony Short was my first manager, and I was asked to implement the ANSI.SYS driver to support up to four programs. This driver interpreted ANSI escape sequences for cursor motion, text bolding, etc., and was heavily influenced by the DEC VT100 terminal design.
At lunch I walked next door to the BURGERMASTER and ordered a burger, fries, and chocolate shake. I remember sitting outside on some concrete steps in the sunshine, eating my lunch, and wondering what the next months and years would bring.
PS: That fall, IBM and Microsoft announced their Joint Development Agreement (to develop OS/2), Microsoft stock split 2-for-1, and Microsoft shipped Windows 1.0. We moved to the brand new Redmond campus in February, 1986, and MSFT went public on 3/13/1986 at $21/share and closed at $27.75/share. Not quite 9 months after I started, my (as yet unvested) stock options were worth $83,250.
PPS: Soon after I started, Microsoft issued us “cardkey” badges, and my number was 80777. I had heard that there were “around 800” employees when I started, so perhaps I was literally the 777th person hired? Grok 3 claims Microsoft 1985 revenues were $140.4 million.
PPPS: I have seen Steve Ballmer, Dave Weil, and Greg Whitten in the the past few years. I am friends with Jo Ann Rahal on Facebook.
Great story, Ben. You may need to explain “telex” and “calling collect” to your younger readers!
Nah…let them Grok(R) it for themselves!
How accurate is Grok?
Very impressive story! Could be google maps has a different route which cuts 50 miles off the route
More likely Google’s digital maps are much more accurate than the maps I used to plan our route 40+ years ago! 😉
Grok (and the other LLMs) are plenty accurate enough to answer questions about 40-year-old technology!